Deposition and decomposition of periodical cicadas (Homoptera: Cicadidae: Magicicada) in woodland aquatic ecosystems
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Deposition and decomposition of periodical cicadas (Homoptera: Cicadidae:
Magicicada) in woodland aquatic ecosystems
Author(s): Corey L. Pray, Weston H. Nowlin, and Michael J. Vanni
Source: Journal of the North American Benthological Society, 28(1):181-195. 2009.
Published By: The Society for Freshwater Science
DOI: http://dx.doi.org/10.1899/08-038.1
URL: http://www.bioone.org/doi/full/10.1899/08-038.1
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.J. N. Am. Benthol. Soc., 2009, 28(1):181–195
Ó 2009 by The North American Benthological Society
DOI: 10.1899/08-038.1
Published online: 23 December 2009
Deposition and decomposition of periodical cicadas
(Homoptera: Cicadidae: Magicicada) in woodland aquatic ecosystems
Corey L. Pray1
Department of Geography, Miami University, 216 Shidler Hall, Oxford, Ohio 45056 USA
Weston H. Nowlin2
Department of Biology, Aquatic Station, Texas State University, San Marcos, Texas 78666 USA
Michael J. Vanni3
Department of Zoology, Miami University, 212 Pearson Hall, Oxford, Ohio 45056 USA
Abstract. Many freshwater ecosystems receive allochthonous resource subsidies from adjacent terrestrial
environments. In eastern North American forests, geographic broods of periodical cicadas emerge every 13
to 17 y to breed, and local abundances can sometimes be .300 individuals/m2. Most individuals avoid
predation, senesce after breeding, and become a resource pulse for forest ecosystems; some cicada carcasses
enter freshwater ecosystems where they represent a detrital resource pulse. Here, we present a 2-part study
in which we examined the deposition of cicada detritus into woodland ponds and low-order streams in
southwestern Ohio during the emergence of Brood X periodical cicadas. We compared the deposition of
nutrients associated with periodical cicada detritus and terrestrial leaf litter into small woodland ponds and
low-order streams. We used a laboratory experiment to compare patterns of decomposition and nutrient
release of adult periodical cicada carcasses and sycamore leaf litter. Input of periodical cicada detritus to
woodland streams and ponds was a function of local cicada emergence densities. Organic C loading to
woodland aquatic ecosystems from cicada detritus was substantially less than that from terrestrial leaf litter;
however, the higher mass-specific N and P content of cicada material made cicada detritus a relatively
important nutrient input. N and P deposited in cicada detritus represented 0.2 to 61% of the N and 0.3 to
50% of the P deposited into woodland aquatic ecosystems via terrestrial leaf litter. Decomposition
experiments indicated that cicada detritus was of much higher quality than was sycamore leaf litter; female
and male cicada carcasses lost mass at significantly faster rates than sycamore leaves (female k ¼ 0.05/d,
male k ¼0.04/d, sycamore leaf k ¼0.002/d). Release rates of C, N, and P from cicada carcasses were 4, 39,
and 1503 greater, respectively, than release rates from sycamore leaves. Our study indicates that periodical
cicada detritus can represent a substantial allochthonous resource pulse to forested aquatic ecosystems and
that cicada detritus is of substantially higher quality than is terrestrial leaf litter. These results suggest that
deposition and decomposition of periodical cicada detritus can affect the productivity and dynamics of
woodland aquatic ecosystems and that the role of animal-derived resource pulses to ecosystems requires
further exploration.
Key words: allochthonous subsidy, resource pulse, decomposition, stoichiometry, periodical cicada, nu-
trient recycling, leaf litter.
Virtually all freshwater ecosystems, including
streams, rivers, wetlands, and lakes, receive prey,
1
Present address: Department of Biology, Aquatic nutrients, and energy from the terrestrial environment
Station, Texas State University, San Marcos, Texas 78666 (Wallace et al. 1997, Wipfli 1997, Pace et al. 2004,
USA. E-mail: cp1274@txstate.edu
2
To whom correspondence should be addressed. E-mail: Mehner et al. 2005, Graham et al. 2006). Terrestrially
wnowlin@txstate.edu derived (allochthonous) resource subsidies can pro-
3
E-mail address: vannimj@muohio.edu foundly affect foodweb structure, productivity, and
181182 C. L. PRAY ET AL. [Volume 28 stability of aquatic ecosystems (Wallace et al. 1997, lin et al. 2007) and stimulate community metabolism Huxel and McCann 1998, Huxel et al. 2002, Takimoto (Menninger et al. in press). Despite the implications of et al. 2002, Nowlin et al. 2007). Many terrestrial this large and infrequent resource pulse to the resource subsidies occur as pulsed inputs, either productivity and dynamics of woodland aquatic seasonally (i.e., deposition of terrestrial insects or systems, little is known about its importance relative autumnal leaf litter) or coincident with large pertur- to other resource inputs, such as terrestrial leaf litter. bation events (i.e., large storm event and watershed Terrestrial primary producer detritus is often the runoff; Nowlin et al. 2008). Small-sized aquatic dominant energy and nutrient source in forest ponds ecosystems, such as low-order streams and woodland and streams, but differences in the seasonal timing and ponds, are closely tied to the terrestrial landscape and quality of terrestrial primary producer litter and cicada receive pulses of terrestrial organic matter that affect detritus might lead to substantially divergent commu- community and ecosystem structure and function nity and ecosystem responses to these resource pulses. (Wallace et al. 1997, Rubbo and Kiesecker 2004, Rubbo Leaf litter typically is imported to forest aquatic et al. 2006, Nowlin et al. 2007, 2008). In particular, ecosystems during a cool period of the year (autumn ponds and headwater streams in forested areas receive and early winter) when microbial decomposers and considerable inputs of detritus, such as leaf litter and invertebrate consumers have low metabolic and woody debris, from terrestrial primary producers. feeding rates. Furthermore, much of the energy in leaf These inputs serve as the principal energy and nutrient litter is bound into recalcitrant polyphenolic com- source for their food webs (Wallace 1997, Johnson and pounds, such as lignin (Chapin et al. 2002). Given the Wallace 2005, Rubbo et al. 2006). recalcitrant nature of this resource, it can persist for In eastern North American forests, periodical relatively long periods of time and provide extended cicadas (Homoptera:Cicadidae:Magicicada) synchro- support for food webs. Inputs of terrestrial detritus can nously emerge from below ground in early to support aquatic food webs for years (Wallace et al. midsummer every 13 to 17 y, and local abundances 1997, Hall et al. 2000). In contrast, periodical cicada vary between 60 and 300 individuals/m2 (Dybas and carcasses are a labile nutrient and energy source that is Davis 1962, Rodenhouse et al. 1997, Yang 2004). The deposited in aquatic systems during summer. There- highest emergence densities typically occur along fore, when periodical cicadas die and fall into forest edges or in lowland riparian areas (Dybas and woodland aquatic habitats, they might decompose Davis 1962, Williams et al. 1993, Rodenhouse et al. quickly and potentially could release significant 1997). The synchronous emergence of large numbers of amounts of C, N, and P, which can stimulate individuals satiates predators, so only a fraction of an heterotrophic bacteria and algae (Nowlin et al. 2007). emergent cicada brood falls victim to predation. Most The initial quality of leaf litter and cicada carcasses individuals reproduce, die, and become deposited as undoubtedly differs when they enter stream and pond detritus on the landscape (Williams et al. 1993). These habitats, but how the stoichiometry and nutrient carcasses are a large resource pulse for habitats in release rates of these organic matter sources differ forest ecosystems (Yang 2004, 2006, Nowlin et al. 2007, during the decomposition process is yet to be Menninger et al. in press). determined. Periodical cicada carcasses potentially are a high- We examined nutrient dynamics associated with the quality detritus source, and contain relatively large deposition and decomposition of periodical cicadas in amounts of N, protein, and lipids (Brown and forested aquatic habitats. We had 2 goals: 1) to compare Chippendale 1973). Thus, mass deposition of period- deposition of nutrients and organic matter associated ical cicada carcasses during an emergence event with periodical cicada detritus and terrestrial leaf litter represents a high-quality resource pulse for terrestrial in small woodland ponds and low-order streams, and forest communities, and affects soil microorganisms, 2) to determine patterns of decomposition, nutrient herbaceous vegetation, arthropods, birds, and rodents stoichiometry, and nutrient release rates of periodical (Hahus and Smith 1990, Krohne et al. 1991, Yang 2004, cicada carcasses and terrestrial leaf litter in aquatic 2006, Koenig and Liebhold 2005). A portion of the systems. To address the 1st question, we examined the emergent biomass also is deposited in small woodland inputs of cicada litter into forest aquatic habitats during aquatic habitats, where it might be consumed by larger the summer of an emergence event and estimated aquatic organisms (Williams and Simon 1995, Vokoun loading of P, N, and C from cicada detritus. We then 2000). Deposition of periodical cicada detritus in contrasted these loadings to inputs of terrestrial leaf aquatic systems releases substantial amounts of litter in autumn of the same year. We addressed the 2nd dissolved nutrients that elicit rapid population re- question by conducting a replicated laboratory micro- sponses from bacteria, algae, and invertebrates (Now- cosm experiment in which we examined the decom-
2009] CICADA DECOMPOSITION IN AQUATIC ECOSYSTEMS 183
FIG. 1. Geographic distribution of Brood X periodical cicadas that emerged in summer 2004. The shaded grey area is the
geographic distribution of Brood X cicadas. The location of the study area within the state of Ohio is indicated with a star.
position rates of periodical cicada carcasses and six 30-m stream segments within 10 km of Oxford,
terrestrial leaf litter under controlled conditions. Ohio. We chose all pond and stream sites before the
Brood X cicada emergence period began. We chose
Methods sites that were in lowland areas with abundant woody
vegetation. Detailed descriptions of pond and stream
Our study was conducted in southwestern Ohio
sites and the sampling design used to collect cicada
(USA) in summer 2004. Starting in mid-May 2004,
litter are presented in Nowlin et al. (2007). Briefly, we
periodical cicadas emerged over much of Ohio as part of
used plastic litterfall baskets to estimate loadings of
the Brood X emergence (Fig. 1). Brood X cicadas emerge
cicada detritus during the emergence period. We
in the midwestern and eastern US over an area of 15
collected all cicada-derived litter from baskets every
states (Marshall 2001). Cicadas emerging at our study
3 to 5 d and sorted it into 5 specific types: whole
sites were composed of 3 species: Magicicada cassini,
nymphs, molted nymph exoskeletons, whole adults,
Magicicada septendecim, and Magicicada septendecula.
severed adult heads with attached wings, and severed
wings (Williams et al. 1993). We determined dry mass
Deposition of cicada detritus and leaf litter
deposition (g/m2) for each basket, and we estimated
To estimate the deposition of periodical cicada litter areal nutrient loading (mmol/m2 of C, N, and P) in
into woodland aquatic habitats, we studied 10 each individual basket during the emergence period
temporary and semipermanent woodland ponds and by multiplying the total mass of each litter type184 C. L. PRAY ET AL. [Volume 28
deposited in a basket over the emergence period by the those of Killingbeck (1996), who reported %N and %P
amount of C, N, and P in each litter type (see below). values for senesced red maple, hickory, beech, tulip
Detailed methods used to measure nutrient content in tree, and white oak leaves that were 0.83 6 0.24% and
each cicada litter type are presented in Nowlin et al. 0.05 6 0.02% (mean 6 1 SD), respectively. Therefore,
(2007). We determined ecosystem-level areal loadings we calculated mean values for mass-specific leaf
of cicada litter in individual ponds and stream reaches nutrient content from Killingbeck’s (1996) values and
from the mean deposition of cicada matter and our analyses (C: 46.90%, N: 0.877%, P: 0.045%). We
nutrients in all baskets within each pond or stream used these percentages and the dry mass deposition of
reach. Open litterfall baskets are used commonly for leaf litter (g/m2) into each basket to estimate nutrient
estimating deposition of organic matter (OM) into (C, N, and P) loading from leaf litter into baskets at
aquatic systems; however, it is likely that our pond and stream sites. We determined leaf litter
measurements of cicada OM and nutrient deposition loading into each individual basket (mg dry mass/
are underestimates because of losses to scavenging m2 and mmol/m2 of C, N, and P from late September
birds and mammals (Yang 2004, WHN, personal to early December) and estimated ecosystem-level
observation). areal loadings of leaf litter into individual ponds and
We replaced litter baskets in the same locations at all stream reaches by calculating the mean amount of leaf
sites in early September 2004 to estimate deposition of matter and nutrients deposited in all baskets within an
terrestrial leaf litter into the same ponds and streams. individual pond or stream reach.
From late September until early December, we
removed all leaf litter material collected in baskets Cicada emergence densities
every 3 to 4 wk and brought it to the laboratory, where
We also wanted to know whether cicada deposition
we dried it at 608C for 48 to 72 h and determined dry
rates into woodland ponds and low-order streams
mass (g). The dominant tree species and their relative
were related to local cicada emergence densities. We
abundances varied among all sites, but several tree
estimated cicada emergence densities around wood-
species, including sycamore (Platanus occidentalis), red
land ponds and streams by counting cicada emergence
maple (Acer rubrum), hickory (Carya tomentosa), beech
holes in the ground (Williams et al. 1993, Rodenhouse
(Fagus spp.), tulip tree (Liriodendron tulipifera), and
et al. 1997, Whiles et al. 2001). We estimated emergence
white oak (Quercus alba), were common at most sites.
densities in the areas immediately adjacent to all
Leaf litter collected in baskets was composed of
ponds and stream sites ;2 to 3 wk after the onset of
leaves from multiple species, and many leaves were
the Brood X emergence (Williams et al. 1993). Detailed
fragmented and difficult to identify. We used the
description of the rationale and methods used to
nutrient content of sycamore to estimate nutrient
determine cicada emergence densities are presented in
loading associated with leaf litter deposition. Syca-
appendix A in Nowlin et al. (2007). Briefly, at pond
more is an abundant tree species in riparian zones in
sites, we counted cicada emergence holes along
Ohio (Kupfer 1996, Vadas and Sanger 1997). We
transects radiating out in 3 randomly selected compass
collected sycamore leaf litter from a riparian area in
directions from the edge of a pond. At stream sites, we
January 2004, dried 6 batches of 5 to 7 individual
estimated cicada emergence densities along 2 transects
leaves at 608C for 48 h, and homogenized each batch.
that were parallel to the stream banks. We calculated
We analyzed C, N, and P content of batch samples. We
average local cicada emergence density (number/m2)
measured C and N of leaf material with a Perkin Elmer
at each site as the mean emergence density of all
2400 Series II CHN analyzer (Perkin Elmer, Boston,
quadrat counts taken at a site.
Massachusetts). We digested samples with HCl and
measured PO43– on a Lachat QuikChemt FIAþ 8000
Decomposition and nutrient release of cicada detritus
Series autoanalyzer (Lachat Instruments, Loveland,
and leaf litter
Colorado). For comparison purposes, we also obtained
%N and %P (by mass) of senesced leaf litter from We examined patterns of decomposition and nutri-
Killingbeck’s (1996) review. We calculated the average ent release in cicada detritus and terrestrial leaf litter in
%N and %P of senesced leaves of red maple, hickory, laboratory experiments. Use of a laboratory setting to
beech, tulip tree, and white oak from table 1 in examine decomposition and nutrient release elimi-
Killingback (1996). The %C values for these litter types nates factors that can affect OM decomposition, such
were not reported in Killingback (1996). as water movement (Hoover et al. 2006) and the
Our analyses of sycamore leaves indicated that C, N, presence of detritivorous metazoans (Cummins et al.
and P were 46.90, 1.01, and 0.04% of sycamore litter 1989, Dangles and Guerold 2001). Thus, our laboratory
dry mass, respectively. These values are similar to experiments provide insight into the relative quality of2009] CICADA DECOMPOSITION IN AQUATIC ECOSYSTEMS 185 periodical cicadas and leaf litter as detritus sources among pairs of males and females was small, so we and the potential input of nutrients from both OM used these values as estimates of initial dry mass of types to the nutrient cycles of woodland aquatic cicada carcasses. ecosystems. We collected live male and female adult cicadas Data analysis from the Miami University campus (Oxford, Ohio) during the peak of the emergence period. We We compared loadings of cicada matter, leaf litter, examined decomposition of male and female cicadas and associated nutrients between ponds (n ¼ 10) and separately because males and females are likely to streams (n ¼ 6) with 1-way analysis of variance differ in their nutrient content and biochemical (ANOVA). We plotted loadings of cicada detritus and composition; male bodies have a large resonance associated nutrients to sites as a function of the local chamber that is used for calling, and female bodies cicada emergence density, and used ordinary least- are full of eggs and reproductive organs. We used M. squares (OLS) linear regression to examine the cassini adults for cicada decomposition experiments relationship between deposition and the local emer- because this species was the most abundant species at gence density of periodical cicadas. We inferred most field locations (WHN, personal observation). significance at a 0.05. We used essentially the same protocol to examine To assess potential differences in decomposition of decomposition of cicada adults and terrestrial leaf the 3 detritus types in the laboratory experiments, we litter. We placed 2 male or 2 female adult M. cassini plotted % initial dry mass remaining of each detritus cicadas into individual 125-mL high-density polyeth- type (adult female cicada carcasses, adult male cicada ylene (HDPE) bottles (18 replicate bottles/gender carcasses, and sycamore leaf litter) as a function of treatment) and ;1 g of an intact piece of dry sycamore time (d) and fitted an exponential decay model to the leaf litter into 18 individual 125-mL HDPE bottles. We data to produce a decay constant (k) for each detritus collected water from a nearby pond (Miami University type. To determine whether mass loss rates differed Ecological Research Center) in a large plastic carboy, between detritus types, we regressed ln(x)-trans- screened it through 63-lm Nitex mesh to remove formed % initial mass remaining of each detritus type macroinvertebrates, and added 125 mL of the screened as a function of day, and compared slopes of these water to each bottle. We used this water source relationships with analysis of covariance (ANCOVA). because decomposition experiments were conducted Percent initial mass remaining was the dependent in conjunction with an outdoor pond mesocosm variable, detritus type was the independent (categor- experiment that used this water source (Nowlin et al. ical) variable, and day was the covariate. We used a 2007). We inserted foam stoppers loosely into the sequential Bonferroni procedure to adjust a (Rice 1989, bottle tops and incubated bottles in a walk-in Moran 2003) for multiple comparisons (females vs environmental chamber at 218C under constant full- males, females vs leaves, and males vs leaves). We spectrum fluorescent illumination (;100 lmol quanta ranked response-variable p-values from least to great- m2 s1). We removed 3 replicate bottles containing est and compared the lowest p-value to a/j, where j is male and female cicadas from the environmental the number of comparisons (a ¼ 0.05/3 ¼ 0.017). We chamber on days 5, 6, 8, 12, 18, and 20, and 3 replicate inferred significance if the p-value of a response bottles containing sycamore leaf material on days 3, 8, variable was lower than the adjusted a. We compared 10, 12, and 20. On each sampling day, we removed all greater p-values progressively to j 1, j 2, etc., until remaining cicada or leaf material from each bottle and the p-value of a response variable exceeded the dried it at 608C for 48 h to determine dry mass. We adjusted a. homogenized dried material from each bottle and We examined changes in nutrient content of the 3 analyzed C, N, and P content of the material. detritus types during the decomposition experiment We estimated mass loss rates of cicada and sycamore by plotting molar nutrient ratios (C:N, C:P, and N:P) as leaf material during decomposition experiments from functions of day and analyzing with OLS regression. plots of the % initial dry mass remaining as a function We compared rates of change in nutrient ratios of the 3 of time (days). We could not directly determine the detritus types over the course of the experiment initial dry mass of the pairs of male and female cicadas (slopes of the OLS regressions) with ANCOVA. Molar placed in bottles, so we measured the dry mass of nutrient ratio was the dependent variable, detritus additional pairs of male and female adult M. cassini (n type was the independent (categorical) variable, and ¼ 8 for male and female pairs). Mean (61 SD) dry day was the covariate. We log(x)-transformed data masses (g) of female and male M. cassini pairs were before analyses to meet assumptions of linearity and 0.554 6 0.07 and 0.336 6 0.05, respectively. Variation homogeneity of variances associated with ANCOVA.
186 C. L. PRAY ET AL. [Volume 28
TABLE 1. Nutrient content (mmol/g dry mass) of types of ANOVA (n ¼ 3 for each type). If we detected a
periodical cicada detritus and leaf litter. Sycamore leaf litter significant treatment effect, we determined homoge-
values are based on our analyses. The composite leaf values nous groups with post hoc Tukey HSD tests. We set a
are mean C, N, and P content from our sycamore leaf
0.05 and used a sequential Bonferroni procedure to
analyses and data for 6 tree species commonly found in Ohio
adjust a for multiple comparisons.
riparian forests (Killingbeck 1996).
C content N content P content Results
(mmol/g (mmol/g (mmol/g
Detritus type dry mass) dry mass) dry mass) Deposition of cicada detritus and leaf litter
Nymph 42.92 6.61 0.17 All types of cicada detritus and sycamore leaf litter
Exuvium 37.32 6.46 0.02 had similar mass-specific amounts of C (;40 mmol
Adult female body 44.31 8.42 0.25 C/g; Table 1). Cicada detritus had high amounts of N
Adult male body 47.50 7.23 0.22
Head and wing pair 42.17 8.34 0.22 (mean ¼ 7.70 mmol/g; Table 1). Most types of cicada
Wing 41.04 8.99 0.07 detritus had ;0.20 mmol/g P. N and P content were
Sycamore leaf 39.08 0.72 0.001 lower in composite leaf litter (0.63 mmol/g N, 0.02
Composite leaf 39.08 0.63 0.02 mmol/g P) than in cicada detritus.
All pond and all stream sites received inputs of
cicada detritus during the emergence of Brood X
We used a sequential Bonferroni procedure to adjust a
periodical cicadas; however, the amount of cicada
for multiple comparisons.
detritus deposited varied among sites. Aquatic eco-
We used changes in mass and nutrient content of
systems received 0.03 to 16.99 g/m2 (3.27 6 4.66 g/m2;
each detritus type during decomposition experiments
x̄ 6 1 SD) of cicada detritus dry mass. Stream sites
to compare short-term nutrient release rates of each
received significantly larger cicada detritus inputs than
detritus type. We calculated release rate (RR; lmol/d)
did pond sites (ANOVA, F1,15 ¼ 22.55, p , 0.001; Fig.
of C, N, and P from each detritus type as:
2). Across all sites, whole adult cicadas and nymphal
RRðlmol=dÞ ¼ ð½DM0 Nut0 ½DM20 Nut20 Þ=20 exuviae were the most common cicada detritus type
where DM is the dry mass (mg) of items at day 0 and deposited (Fig. 2); the combination of these detritus
day 20 and Nut is the proportional nutrient content (C, types made up 72% and 98% of dry mass deposition in
N, or P mg/mg dry mass) on days 0 and 20 of the pond and stream ecosystems, respectively. Males and
experiment. We could not determine the initial females were deposited in a ;1:1 ratio (WHN,
nutrient content of the leaf and cicada material placed unpublished data).
into bottles. Therefore, we used the nutrient content of
30 female and 30 male adult cicadas and 6 homoge-
nized batches of 5 to 7 sycamore leaves (see above) and
assumed that initial content of female adult cicadas
was 50.83% C, 11.78% N, and 0.77% P; initial content
of adult male cicadas was 52.26% C, 10.12% N, and
0.69% P; and initial content of sycamore leaves was
46.90% C, 1.01% N, and 0.04% P. We determined the
initial dry mass of sycamore leaf litter before incuba-
tion, but we could not do this for female and male
cicadas. Therefore, we assumed that the initial dry
mass of a pair of female cicadas was 554 mg and a pair
of male cicadas was 336 mg on the basis of prior
analyses (n ¼ 8 for both male and female pairs; see
above). We compared release rates of C, N, and P for
each detritus type with 1-way ANOVA (n ¼ 3 for each
detritus type). If we detected a significant treatment
effect, we determined homogenous groups with post FIG. 2. Mean (þ1 SD) dry mass deposition (g/m2) of
hoc Tukey honestly significant difference (HSD) tests. periodical cicada detritus in woodland pond and headwater
We compared molar ratios of nutrients released from stream sites. Ad ¼ adult cicadas, Ex ¼ nymphal exuviae, Ny ¼
cicada carcasses and sycamore leaf litter (C:P, N:P, and whole nymphs, Wi ¼ wings only, HþW ¼ severed heads with
C:N) over the experimental period with 1-way wings attached, Tot ¼ total cicada material.2009] CICADA DECOMPOSITION IN AQUATIC ECOSYSTEMS 187
(mmol/m2)
TABLE 2. Means (61 SD) and ranges of dry mass (g/m2) and nutrient loading (mmol/m2) of leaf litter and periodical cicada detritus to woodland aquatic
0.08 6 0.13
0.64 6 0.97
1.58 61.04
0.002–0.32
0.56–3.60
0.002–3.60
P
56.23 6 35.34
23.44 6 33.41
18.88–122.14
0.31–122.14
(mmol/m2)
3.77 6 4.46
0.31–12.02
N
Cicada detritus
364.04 6 230.79
150.89 6 217.58
22.99 6 29.04
123.30–798.26
1.43–798.26
(mmol/m2)
1.43–77.53
C
0.53 6 0.64
7.85 6 4.92
3.27 6 4.66
2.61–16.99
0.03–16.99
Dry mass
0.03–1.69
(g/m2)
FIG. 3. Relationship between local emergence density of
periodical cicadas (estimated from emergence holes) and dry
mass deposition of periodical cicada detritus in woodland
aquatic ecosystems. The regression line was generated using
(mmol/m2)
6.00 6 2.13
5.78 6 2.10
5.92 6 2.05
3.38–8.95
2.60–8.19
2.60–8.95
all sites. See Results for the separate results of ordinary least-
squares regressions for pond sites and stream sites.
P
Loadings of C, N, and P from cicada detritus into
aquatic habitats followed the same pattern as dry mass
189.00 6 67.04
181.92 6 66.22
186.35 6 64.58
106.63–281.82
81.77–258.06
81.77–281.82
deposition. On average, stream sites received signifi-
(mmol/m2)
cantly greater loadings of C, N, and P from cicada
detritus than did pond sites (C: F1,15 ¼ 22.23, p , 0.001;
N
N: F1,15 ¼ 22.50, p , 0.001; P: F1,15 ¼ 21.01, p , 0.001;
Table 2). Across all pond and stream sites, the C, N,
and P loadings from cicada detritus were 150.89
Leaf litter
mmol/m2, 23.44 mmol/m2, and 0.64 mmol/m2,
11724.10 6 4158.38
11284.70 6 4108.04
11559.32 6 4006.10
6614.22–17482.02
5072.46–16007.84
5072.46–17482.02
respectively (Table 2).
(mmol/m2)
ecosystems within 10 km of Oxford, Ohio, USA.
Differences in cicada detritus dry mass, C, N, and P
loadings between stream and pond sites were associ-
C
ated with differences in the local emergence density of
periodical cicadas at the 2 site types. Dry mass
deposition of cicada detritus was a positive function
of local emergence density in the immediate area
around aquatic habitats (y ¼ 0.141x þ 0.073, r2 ¼ 0.75,
300.00 6 106.41
288.76 6 105.12
295.79 6 102.51
169.25–447.34
129.80–409.62
129.80–447.34
F1,15 ¼ 42.51, p , 0.001; Fig. 3). Stream sites typically
Dry mass
(g/m2)
had higher emergence densities (42.75 6 39.67) than
did pond sites (10.66 6 8.19). When pond sites were
considered alone, the relationship between emergence
density and cicada litter deposition was not significant
(F1,9 ¼ 1.71, p ¼ 0.228). However, when stream sites
were considered alone, the relationship between
Stream
Sites
Pond
emergence density and cicada litter deposition was
All
significant (F1,5 ¼ 11.19, p ¼ 0.029).188 C. L. PRAY ET AL. [Volume 28
Deposition of leaf litter did not differ significantly
between pond and stream sites (F1,15 ¼ 0.32, p ¼ 0.581;
Table 2). Dry mass deposition of leaf litter in ponds and
streams ranged from 129.80 to 447.34 g/m2. Leaf litter
inputs provided substantial C loadings (5072.46–
17,482.02 mmol/m 2 ), but loadings of N and P
associated with leaf litter were considerably smaller
and ranged from 81.77 to 281.82 mmol/m2 N and 2.60
to 8.95 mmol/m2 P. Even though the dry mass and C
loading of terrestrial leaf litter were much larger than
those of cicada detritus (cicada litter dry mass
represented 0.01–5.42% of the dry mass deposition of
leaf litter), the much higher mass-specific N and P
content of cicada detritus made it a relatively important
nutrient input in some forest aquatic ecosystems,
especially stream sites. On the basis of deposition data
FIG. 4. Mean (61 SE) % initial dry mass remaining of
presented in Table 2, N and P deposited into ponds via adult male and female periodical cicada carcasses and
cicada litter represented 0.16 to 6.89% of the N (x̄ ¼ sycamore leaf litter in laboratory decomposition experi-
2.23%) and 0.3 to 5.76 % of the P (x̄ ¼ 1.35%) deposited ments. See Methods for calculation of decomposition
in autumnal leaf litter. In contrast, in streams, N and P constants (k) for each detritus type.
deposited in cicada litter represented 10.94 to 61.18% (x̄
¼ 33.40) of the N and 10.26 to 50.36% (x̄ ¼ 29.05%) of the (ANCOVA: F1,41 ¼ 15.75, p , 0.001), but C:P of both
P deposited in autumnal leaf litter. female (ANCOVA: F1,41 ¼ 351.77, p , 0.001) and male
(ANCOVA: F1,41 ¼ 554.81, p , 0.001) cicada carcasses
Decomposition and nutrient release of cicada detritus changed more slowly than did C:P of sycamore leaf
and leaf litter litter (Fig. 5B). N:P of all detritus types increased
Adult female and male cicada carcasses lost mass markedly as decomposition progressed (Fig. 5C). Rates
quickly during decomposition experiments, and the of N:P change did not differ significantly between
relationships between time and % inital mass remain- male and female cicada carcasses (ANCOVA: F1,41 ¼
ing were best described by exponential decay func- 0.77, p ¼ 0.385), but N:P of both female (ANCOVA:
tions (Fig. 4). Mass loss rates did not differ F1,41 ¼ 40.83, p , 0.001) and male (ANCOVA: F1,41 ¼
significantly between male and female cicada carcasses 45.88, p , 0.001) cicada carcasses changed faster than
(female: k ¼0.051, male: k ¼0.037; ANCOVA: F1,41 ¼ did N:P of sycamore leaf litter.
1.58, p ¼ 0.216). Mass loss rates for female (ANCOVA: Differences in rates of mass loss and nutrient
F1,38 ¼ 57.59, p , 0.001) and male (ANCOVA: F1,38 ¼ stoichiometry among detritus types during decompo-
61.80, p , 0.001) cicada carcasses were significantly sition led to contrasting nutrient release rates (Fig. 6A–
greater than those for sycamore leaf litter (sycamore C). Female and male cicada carcasses released signif-
leaf: k ¼ 0.002). icantly greater amounts of C, N, and P than did
C:N, C:P, and N:P of adult male and female cicada
sycamore leaf litter (ANOVA: C: F1,8 ¼ 6.96, p ¼ 0.027;
carcasses and sycamore leaf litter changed significantly
N: F1,8 ¼ 28.67, p ¼ 0.001; P: F1,8 ¼ 44.98, p , 0.001).
(except C:P of sycamore leaf litter) as they decomposed
Release rates of C, N, and P were 4, 39, and 1503
(Fig. 5A–C; Table 3). C:N increased faster in female
greater, respectively, from cicada carcasses than from
than in male cicada carcasses (ANCOVA: F1,41 ¼ 34.28,
sycamore leaf litter. In fact, the negative N release rate
p , 0.001; Fig. 5A). Patterns of change in C:N of both
female (ANCOVA: F1,41 ¼ 569.82, p , 0.001) and male of sycamore leaf litter (Fig. 6B) indicated that N was
(ANCOVA: F1,41 ¼ 1050.59, p , 0.001) cicada carcasses taken up during the experiment. In general, nutrient
were dramatically different from patterns of change in release rates of female cicada carcasses were greater
C:N of sycamore leaf litter (Fig. 5A). The slope of the than those of males, but this difference was significant
relationship between sycamore leaf litter C:N and day only for P (Fig. 6C). On the basis of nutrient content of
was negative, indicating a decrease in C:N with time, cicada carcasses and nutrient release rates observed in
whereas the slope of this relationship was positive for decomposition experiments, female and male cicada
both female and male cicada carcasses (Table 3). C:P carcasses released 64%, 73%, and 91% of the C, N, and
increased faster in female than in male cicada carcasses P they contained over the 20-d period. In contrast,2009] CICADA DECOMPOSITION IN AQUATIC ECOSYSTEMS 189
sycamore leaf litter released 7%, 16%, and 54% of the
C, N, and P it contained over the 20-d period.
C:P of nutrients released by female and male cicada
carcasses did not differ significantly ( p ¼ 0.999), but
C:P of nutrients released by male and female cicada
carcasses was significantly lower (47:1) than that of
nutrients released by sycamore leaf litter (1717:1)
(ANOVA: F2,8 ¼ 26.78, p ¼ 0.001; female cicada vs
male cicada: p , 0.999; female cicada vs sycamore: p ¼
0.002; male cicada vs sycamore: p ¼ 0.002; Fig. 6D).
Sycamore leaf litter showed a net increase in N content
(e.g., a negative N release rate); thus C:N and N:P
ratios for released nutrients were not calculated for
sycamore leaf litter because the values would be
negative. The C:N and N:P ratios for nutrients released
by female and male cicada carcasses were not
significantly different (ANOVA: C:N: F1,5 ¼ 0.93, p ¼
0.776; N:P: F1,5 ¼ 0.92, p ¼ 0.776; Fig. 6E, F).
We used our estimates of nutrient loss from cicada
carcasses and sycamore leaf litter to calculate short-
term C, N, and P release by these detritus sources in
woodland aquatic ecosystems. We assumed that all
types of cicada material (i.e., bodies, exuvia, head/
wing pairs) had the same patterns of % nutrient release
as adult cicada carcasses and that all types of terrestrial
leaf litter had the same % nutrient release as sycamore
leaf litter over a 20-d period (i.e., same percentage of
total C, N, and P contained in detritus was released
over 20 d). We coupled these nutrient release estimates
with measurements of C, N, and P deposition of cicada
and leaf detritus into woodland ponds and streams
(Table 2) to determine the amount of C, N, and P
released in woodland ponds and streams during a 20-
d period. These nutrient release estimates represent the
amount of nutrients released by cicada detritus and
leaf litter into the surrounding water across the range
of cicada detritus and leaf litter observed in woodland
ponds and streams. Across all pond and stream
ecosystems, cicada litter released 96.57 mmol C (range
¼ 0.92–510.89 mmol), 17.70 mmol N (range ¼ 0.23–
92.22 mmol), and 0.60 mmol P (range ¼ 0.002–3.32
mmol). In these same ecosystems, terrestrial leaf litter
released 809.15 mmol C (range ¼ 355.04–1223.74
mmol) and 3.16 mmol P (range ¼ 1.40–4.83 mmol).
We estimate that leaf litter would absorb 29.82 mmol N
(range ¼ 13.08–45.00 mmol) over a 20-d period. Thus,
the potential importance of periodical cicada detritus
as a source of N and P to aquatic communities is more
apparent when field OM deposition and laboratory
FIG. 5. Mean (61 SE) molar C:N (A), C:P (B), and N:P (C) decomposition data are coupled.
of adult female and male periodical cicada carcasses and
sycamore leaf litter during laboratory decomposition exper-
iments. Results of ordinary least-squares (OLS) regression untransformed, whereas the OLS regression analyses in
analyses are presented in Table 3. Data presented here are Table 3 were generated with log(x)-transformed data.190 C. L. PRAY ET AL. [Volume 28
TABLE 3. Results of ordinary least-squares regression analyses for changes in nutrient ratios of adult female and male cicada
carcasses and sycamore leaf litter during laboratory decomposition experiments. Analyses presented here were generated with
log(x)-transformed data. Untransformed data are presented in Fig. 5.
Nutrient ratio (molar) Detritus type Equation r2 F1,20 p
C:N Female cicada carcass y ¼ 0.012x þ 0.80 0.70 18.09 ,0.001
Male cicada carcass y ¼ 0.006x þ 0.74 0.20 9.64 0.006
Sycamore leaves y ¼ 0.008x þ 1.75 0.43 8.68 0.008
C:P Female cicada carcass y ¼ 0.039x þ 2.33 0.90 72.54 ,0.001
Male cicada carcass y ¼ 0.035x þ 2.20 0.92 66.10 ,0.001
Sycamore leaves y ¼ 0.0079x þ 3.58 0.19 3.20 0.089
N:P Female cicada carcass y ¼ 0.029x þ 1.48 0.89 62.45 ,0.001
Male cicada carcass y ¼ 0.027x þ 1.53 0.91 47.92 ,0.001
Sycamore leaves y ¼ 0.016x þ 1.83 0.59 21.66 ,0.001
Discussion the pond sites were in wood lots adjacent to
agricultural fields. Disturbance regimes of these
Deposition of cicada detritus and terrestrial leaf litter
respective environments might affect the number of
Detritus generated from the emergence of Brood X cicadas emerging from each environment and the
periodical cicadas provided a substantial pulse of number of cicadas selecting these sites for oviposition.
allochthonous resources to aquatic ecosystems in Dry mass deposition of periodical cicada detritus in
forested areas. Both woodland ponds and low-order summer represented a fraction of the dry mass loading
streams received a pulse of periodical cicada detritus; of terrestrial leaf litter into aquatic systems in autumn
however, the size of this input was a function of the (;1.23% of dry mass deposition of leaf litter).
local emergence density of the cicadas. The range in Likewise, loading of C from periodical cicadas was
local emergence densities across all woodland aquatic 1.44% of loading of C from leaf litter. Leaf litter and
habitats spanned almost 2 orders of magnitude (2–99 cicada litter contained approximately the same amount
individuals/m2), but stream sites had significantly of C per unit mass (;40 mmol/g dry mass), but the
higher emergence densities than did pond sites. greater total mass deposition of leaf litter resulted in
Periodical cicada emergence densities around our much greater C loadings associated with leaf litter
woodland aquatic systems are within the range of than with cicada litter. To examine these allochthonous
those values reported by others (from 2 to .350 C inputs in the context of autochthonous C production
individuals/m2) (Dybas and Davis 1962, Williams et in woodland aquatic ecosystems, we compared C
al. 1993, Rodenhouse et al. 1997, Whiles et al. 2001, loadings (mmol C m2 d1) from leaf litter and
Yang 2004). periodical cicada litter to in situ measurements of
The systematic differences between stream and primary production in woodland aquatic habitats.
pond sites in periodical cicada emergence densities Roberts et al. (2007) estimated that the daily gross
and deposition rates might be related to habitat primary production in Walker’s Branch, a forested
preferences of breeding and ovipositing periodical headwater stream in Tennessee, was ;43 mmol C m2
cicadas. In our study, average emergence densities d1 (i.e., annual total production divided by 365 d).
were ;43 higher at stream sites than at the pond sites. Daily allochthonous leaf litter input to our forested
Cicadas might prefer lowland riparian habitats to streams in Ohio was 30.92 mmol C m2 d1, and
upland forests (Dybas and Davis 1962, Williams et al. annual cicada input was 1.0 mmol C m2 d1 (i.e., C
1993, Whiles et al. 2001), and cicadas prefer forest loading from each detritus type in Table 2 divided by
edges to interiors (Lloyd and White 1976, Lloyd and 365 d). Thus, in some woodland aquatic systems,
Karban 1983, Rodenhouse et al. 1997). Many of the allochthonous C inputs from leaf litter is of similar
forest pond sites in our study were in the interior of magnitude to autochthonous primary production, but
wood lots and had almost complete canopy cover over C inputs via cicada detritus is a fraction of both leaf
pond surfaces. In contrast, all stream sites had open litter and in situ primary production.
canopy over portions of the streambed and were In contrast to the relatively small C inputs associ-
entirely within lowland habitats. Indeed, stream banks ated with cicada detritus, its higher N and P content
represent edge habitat that periodical cicadas might made cicada detritus a sizable N and P pulse in some
select when breeding and ovipositing. In addition, all aquatic habitats, particularly stream ecosystems with
stream sites were within public natural lands, whereas high cicada emergence densities. At stream sites, the2009] CICADA DECOMPOSITION IN AQUATIC ECOSYSTEMS 191 FIG. 6. Mean (þ1 SE) release rates (lmol/d) of C (A), N (B), and P (C) and the molar C:P (D), C:N (E), and N:P (F) of nutrients released from adult female and male periodical cicada carcasses and sycamore leaf litter during laboratory decomposition experiments. Homogeneous treatment groups (Tukey tests) share a letter. maximum observed amount of N and P deposited in and streams in the same area ranged from 0.31 to cicada detritus represented 61% of the N and 50% of 122.14 mmol m2 y1 (x̄ ¼ 23.44 mmol m2 y1). These the P associated with the deposition of terrestrial leaf data indicate that the short-term pulsed input of N litter. We also compared nutrient loading from cicada from cicada detritus is of similar magnitude to the detritus to loadings from other outside sources that annual atmospheric inorganic N load to these ecosys- can contribute substantial amounts of nutrients to tems (e.g., cicada N deposition was 1–220% of annual woodland aquatic ecosystems. Atmospheric deposi- atmospheric N deposition). tion of N into aquatic ecosystems is of concern in Small woodland ponds and streams are highly many parts of eastern North America, including subsidized by allochthonous leaf litter. However, our southwestern Ohio (Fenn et al. 1998). We obtained data indicate that N and P associated with periodical data for atmospheric deposition of inorganic N from cicada detritus can become a sizeable pulse (e.g., the National Atmospheric Deposition Program .10% of the N and P associated with autumnal leaf (NADP) website (http://nadp.sws.uiuc.edu/) for a litter deposition) when local emergence densities site at the Miami University Ecology Research Center, exceed ;12 individuals/m2 (Table 2, Fig. 3). Deposi- Oxford, Ohio (NADP monitoring location OH09; also tion of periodical cicada carcasses on the forest floor designated by the US Environmental Protection represented ,1% of the nutrient flux associated with Agency as Clean Air Status and Trends Network site annual leaf litter deposition in an Arkansas forest with OXF 122). Annual atmospheric deposition of inorganic relatively low periodical cicada emergence densities N from 1985 to 2005 in our study area ranged from (;6 individuals/m2) (Wheeler et al. 1992). Emergence 29.21 to 55.57 mmol m2 y1, with a mean annual densities did not exceed 99 individuals/m2 in our deposition of 40.44 mmol m2 y1. During summer study, but others have reported much greater emer- 2004, cicada detritus N loading into woodland ponds gence densities (.350 individuals/m2) (Dybas and
192 C. L. PRAY ET AL. [Volume 28
Davis 1962, Yang 2004). Therefore, N and P loading might have been caused by differences in the
from deposition of periodical cicada detritus probably biochemical composition of male and female cicadas.
exceeds N and P loading from terrestrial leaf litter in Male and female cicadas have approximately the same
some systems. mass-specific C, N, and P content (Table 1), but female
cicadas contain roughly 23 as much soluble protein
Decomposition and nutrient release of cicada detritus and 33 as much lipid as males (Brown and Chippen-
and leaf litter dale 1973). Female cicada carcasses might have faster
nutrient release rates than males because their greater
Nutrient ratios indicated that cicada detritus (C:N ¼
protein and lipid content is more readily released as
6:1, C:P ¼ 173:1) was a much higher-quality resource
they decompose.
than was sycamore leaf litter (C:N ¼ 51:1, C:P ¼ 2332:1)
to aquatic microbes and metazoan detritivores (Sterner
Implications of periodical cicada detritus for woodland
and Elser 2002, Cross et al. 2005). Potential differences
aquatic ecosystems
in quality and lability were reflected in mass loss in
decomposition experiments. Cicada carcasses decom- Our data indicate that the amount of N and P
posed at a significantly faster rate than did leaf litter entering woodland streams and ponds via cicada
(;30% of cicada carcass initial mass remaining at the detritus in years of emergence (every 17 y) can be a
end of the experiment vs ;90% of initial sycamore leaf sizable pulsed input when compared with other
litter mass remaining). In leaf litter, C associated with allochthonous resource subsidies that occur within
lignin and cellulose is slow to be released because the same year (i.e., leaf litter and atmospheric
these molecules yield little net energy gain for deposition). Given the relative size, quality, and
microbes during decomposition (Chapin et al. 2002). lability of periodical cicada detritus, it is likely that
C contained in cicada carcasses is in the form of this resource pulse can have large effects on ecosystem
soluble proteins, triglycerides, and chitin (Brown and productivity and community dynamics of forest
Chippendale 1973), and these labile fractions are aquatic ecosystems. Indeed, deposition of periodical
quickly leached out of cicada bodies. cicada carcasses into small-order streams can lead to
Patterns of nutrient stoichiometry and nutrient rapid, short-term increases in daily whole-stream
release during decomposition differed substantially respiration rates (Menninger et al., in press). Deposi-
between cicada carcasses and sycamore leaf litter. tion of periodical cicada carcasses to woodland pond
Release rates of C, N, and P were considerably greater mesocosms at a range of commonly observed emer-
from cicada carcasses than from sycamore leaf litter. gence densities leads to rapid increases in total and
Thus, C:N, C:P, and N:P of cicada carcasses increased dissolved nutrients that quickly led to increases in
significantly as decomposition progressed. In contrast, biomass of bacteria, phytoplankton, zooplankton,
C:N of sycamore leaf litter decreased during decom- periphyton, and snails (Nowlin et al. 2007).
position as leaf litter gained N content. The net gain of The amount of cicada detritus entering an individ-
N content was probably a consequence of microbial ual forest aquatic ecosystem and the magnitude of
colonization and increasing biomass on leaf surfaces effects on community and ecosystem processes is a
during the experimental period (Cross et al. 2005) and function of the local cicada emergence density (Nowlin
indicates that the microbial communities on leaf litter et al. 2007, Menninger et al., in press; this study). In
surfaces took up dissolved N from the external addition, environmental conditions in local ecosystems
environment. Cicada carcasses released nutrient at have the potential to affect adjacent or downstream
much lower C:P (47:1) than did sycamore leaf litter aquatic systems, so that habitats with low emergence
(1717:1). These differences in the ratios at which densities might be affected by the deposition of large
nutrients were released from cicada carcasses and leaf amounts of periodical cicada detritus in another
litter indicate the higher quality of cicada carcasses as a system. Individual reaches can retain a large fraction
source of OM. of deposited cicada detritus (Menninger et al., in
The most striking contrasts observed in our decom- press), but streams export a substantial proportion of
position experiments were the differences between energy and nutrients from animal-derived resource
cicada carcasses and leaf litter, but the nutrient pulses to downstream systems because of water
dynamics of decomposing male and female cicada movement (e.g., Mitchell and Lamberti 2005). A large
carcasses also differed. C:N and C:P increased faster in proportion of the nutrients released by cicada pulses
female than in male carcasses during decomposition might be exported downstream and used in reaches
because of generally faster female nutrient release with low local densities of emergent cicadas. Thus,
rates. Faster nutrient release rates by female cicadas deposition of periodical cicada detritus into aquatic2009] CICADA DECOMPOSITION IN AQUATIC ECOSYSTEMS 193
systems in forested landscapes can be extremely 2001, Naiman et al. 2002, Regester and Whiles 2006).
patchy, but the effects might be far reaching because Resource pulses generated from animal material in
of rapid decomposition of cicada detritus and distri- aquatic ecosystems have received less attention from
bution of nutrients by flow in these ecosystems. ecologists than have pulses of plant material, but our
The effects of deposition of leaf litter in forested study and others indicate that animal-derived al-
aquatic ecosystems probably are less spatially variable lochthonous subsidies can be critical resource pulses
than are the effects of deposition of cicada carcasses. that can alter the dynamics of communities and
Deposition of leaf litter ranged from 129.8 to 447.3 ecosystems.
g/m2 (33 range), whereas deposition of periodical
cicada detritus ranged from 0.03 to 17.0 g/m2 (5703 Acknowledgements
range) at our study sites. Deposited leaf litter can be
retained in stream reaches (Brookshire and Dwire We thank Dan Wannamacher, Lynette Pauly, Jon
2003), but smaller spatial variation in leaf litter inputs, Valente, Giselle Balaguer, and Jim Stoekel for field and
relatively lower nutrient content of leaves, and slow laboratory assistance. Kim Medley, Hank Stevens,
decomposition rates of leaf material suggests that the Marı́a J. González, Matthew Fields, and Tom Crist
effects of leaf litter inputs within an ecosystem should provided advice and intellectual support. Luciana
be more spatially homogeneous than those of cicada Carneiro, Beth Dickman, Janelle Duncan, Marty
detritus and that local effects are less likely to Horgan, Lesley Knoll, Alisa Abuzeineh, Brad Caston,
propagate to other habitats. Alex Smith, Crystal LeBeouf, Thibault Datry, Pamela
Rapid decomposition of periodical cicada detritus Silver, and 2 anonymous referees provided comments
and subsequent nutrient release produce rapid re- that substantially improved this manuscript. Our
sponses in aquatic ecosystems, and both community- research was supported through a National Science
(Nowlin et al. 2007) and ecosystem-level (Menninger et Foundation–Small Grants for Exploratory Research
al., in press) processes respond within days of grant (DEB-0420593) and the Miami University Ecol-
deposition of periodical cicada detritus. However, the ogy of Human-Dominated Landscapes Research Ex-
long-term implications of this resource pulse remain periences for Undergraduates program (DBI-0353915).
unknown. In southwestern Ohio forests, cicada detri-
tus is deposited into aquatic ecosystems every 17 y and Literature Cited
community- and ecosystem-level effects are unlikely to BROOKSHIRE, E. N. J., AND K. A. DWIRE. 2003. Controls on
span this time interval. However, if populations of patterns of course particulate retention in headwater
long-lived consumers (i.e, those that live for .1 y) or streams. Journal of the North American Benthological
organisms that produce persistent resting stages or Society 22:17–34.
seeds respond to pulses of periodical cicada detritus, BROWN, J. J., AND G. M. CHIPPENDALE. 1973. Nature and fate of
then the effects of an emergence event might persist for the nutrient reserves of the periodical (17 year) cicada.
longer than a single growing season. In contrast, leaf Journal of Insect Physiology 19:607–614.
CARLTON, R. G., AND C. R. GOLDMAN. 1984. Effects of a massive
litter is deposited every year, slow to decompose, and
swarm of ants on ammonium concentrations in a
can support aquatic food webs for extended periods
subalpine lake. Hydrobiologia 111:113–117.
(Wallace et al. 1997). Thus, cicada detritus is a much CHAPIN, F. S., P. A. MATSON, AND H. A. MOONEY. 2002.
more pulsed allochthonous resource input than is Principals of terrestrial ecosystem ecology. Springer-
terrestrial leaf litter (Nowlin et al. 2008). Verlag, New York.
Allochthonous resource subsidies can influence the CROSS, W. F., J. P. BENSTEAD, P. C. FROST, AND S. A. THOMAS.
productivity and stability of ecosystems (Polis et al. 2005. Ecological stoichiometry in freshwater benthic
1997, Huxel and McCann 1998, Huxel et al. 2002), and systems: recent progress and perspectives. Freshwater
the quality of these allochthonous resource subsidies Biology 50:1895–1912.
can have fundamental influences on ecological dy- CUMMINS, K. W., M. A. WILZBACH, D. M. GATES, J. B. PERRY, AND
namics (Huxel 1999, Rubbo and Kiesecker 2004). W.B. TALIFERRO. 1989. Shredders and riparian vegetation-
Deposition of periodical cicada carcasses into wood- leaf litter that falls into streams influences communities
of stream invertebrates. BioScience 39:24–30.
land aquatic systems is an allochthonous resource
DANGLES, O., AND F. GUEROLD. 2001. Linking shredders and
pulse that can have profound effects on the produc- leaf litter processing: insights from an acidic stream
tivity and dynamics of woodland aquatic ecosystems. study. International Review of Hydrobiology 86:395–406.
Indeed, allochthonous subsidies of carcasses from DYBAS, H. S., AND D. D. DAVIS. 1962. A population census of
many animal taxa (i.e., ants, salamanders, and fish) seventeen-year periodical cicadas (Homoptera: Cicadiae:
are resource pulses that influence ecological dynamics Magiciada). Ecology 43:432–444.
(Carlton and Goldman 1984, Helfield and Naiman FENN, M. E., M. A. POTH, J. D. ABER, J. S. BARON, B. T. BORMANN,194 C. L. PRAY ET AL. [Volume 28
D. W. JOHNSON, A. D. LEMLY, S. G. MCNULTY, D. E. RYAN, geographic stability of brood distributions. Annals of
AND R. STOTTLEMYER. 1998. Nitrogen excess in North the Entomological Society of America 94:386–399.
American ecosystems: predisposing factors, ecosystem MEHNER, T., J. IHLAU, AND H. DORNER. 2005. Can feeding of fish
responses, and management strategies. Ecological Ap- on terrestrial insects subsidize the nutrient pool of lakes?
plications 8:706–733. Limnology and Oceanography 50:2022–2031.
GRAHAM, M. D., R. D. VINEBROOKE, AND M. TURNER. 2006. MENNINGER, H. L., M. A. PALMER, L. S. CRAIG, AND D. C.
Coupling of boreal forests and lakes: effects of conifer ROCHARDSON. Periodical cicada detritus impacts stream
pollen on littoral communities. Limnology and Ocean- ecosystem function. Ecosystems (in press).
ography 51:1524–1529. MITCHELL, N. L., AND G. A. LAMBERTI. 2005. Responses in
HAHUS, S. C., AND K. G. SMITH. 1990. Food habits of Blarina, dissolved nutrients and epilithon abundance to spawn-
Peromyscus, and Microtus in relation to an emergence of ing salmon in southeast Alaska streams. Limnology and
periodical cicadas (Magicicada). Journal of Mammalogy Oceanography 50:217–227.
71:249–252. MORAN, M. D. 2003. Arguments for rejecting the sequential
HALL, R. O., J. B. WALLACE, AND S. L. EGGERT. 2000. Organic Bonferroni in ecological studies. Oikos 100:403–405.
matter flow in stream food webs with reduced detrital NAIMAN, R. J., R. E. BILBY, D. E. SCHINDLER, AND J. M. HELFIELD.
resource base. Ecology 81:3445–3463. 2002. Pacific salmon, nutrients, and the dynamics of
HELFIELD, J. M., AND R. J. NAIMAN. 2001. Effects of salmon- freshwater and riparian ecosystems. Ecosystems 5:399–
derived nitrogen on riparian forest growth and implica- 417.
tions for stream habitat. Ecology 82:2403–2409. NOWLIN, W. H., M. J. GONZALEZ, M. J. VANNI, M. H. H. STEVENS,
HOOVER, T. M., J. S. RICHARDSON, AND N. YONEMITSU. 2006. M. W. FIELDS, AND J. J. VALENTE. 2007. Allochthonous
Flow-substrate interactions create and mediate leaf litter subsidy of periodical cicadas affects the dynamics and
resource patches in streams. Freshwater Biology 51:435– stability of woodland pond communities. Ecology 88:
447. 2174–2186.
HUXEL, G. R. 1999. On the influence of food quality in NOWLIN, W. H., M. J. VANNI, AND L. H. YANG. 2008. Comparing
resource pulses in aquatic and terrestrial ecosystems.
consumer-resource interactions. Ecology Letters 2:256–
Ecology 89:647–659.
261.
PACE, M. L., J. J. COLE, S. R. CARPENTER, J. F. KITCHELL, J. R.
HUXEL, G. R., AND K. MCCANN. 1998. Food web stability: the
HODGSON, M. C. VAN DE BOGERT, D. L. BADE, E. S. KRITBERG,
influence of trophic flows across habitats. American
AND D. BASTVIKEN. 2004. Whole-lake carbon-13 additions
Naturalist 152:460–469.
reveal terrestrial support of aquatic food webs. Nature
HUXEL, G. R., K. MCCANN, AND G. A. POLIS. 2002. Effects of
427:240–243.
partitioning allochthonous and autochthonous resources
POLIS, G. A., W. B. ANDERSON, AND R. D. HOLT. 1997. Toward an
on food web stability. Ecological Research 17:419–432.
integration of landscape and food web ecology. Annual
JOHNSON, B. R., AND J. B. WALLACE. 2005. Bottom-up limitation
Review of Ecology and Systematics 28:289–316.
of a salamander in a detritus-based food web. Canadian
REGESTER, K. J., AND M. R. WHILES. 2006. Decomposition rates
Journal of Fisheries and Aquatic Sciences 62:301–311. of salamander (Ambystoma maculatum) life stages and
KILLINGBECK, K. T. 1996. Nutrients in senesced leaves: keys to associated energy and nutrient fluxes in ponds and
the search for potential resorption and resorption adjacent forest in southern Illinois. Copeia 4:640–649.
proficiency. Ecology 77:1716–1727. RICE, W. R. 1989. Analyzing tables of statistical tests.
KOENIG, W. D., AND A. M. LIEBHOLD. 2005. Effects of periodical Evolution 43:223–225.
cicada emergence on abundance and synchrony of avian ROBERTS, B. J., P. J. MULHOLLAND, AND W. R. HALL. 2007.
populations. Ecology 86:1873–1882. Multiple scales of temporal variability in ecosystem
KROHNE, D. T., T. J. COUILLARD, AND J. C. RIDDLE. 1991. metabolism rates: results from 2 years of continuous
Population responses of Peromyscus leucopus and Blarina monitoring in a forested headwater stream. Ecosystems
brevicauda to emergence of periodical cicadas. American 10:588–606.
Midland Naturalist 126:317–321. RODENHOUSE, L. N., P. J. BOHLEN, AND G. W. BARRETT. 1997.
KUPFER, A. 1996. Patterns and determinants of edge vegeta- Effects of woodland shape on the spatial distribution and
tion of a midwestern forest preserve. Physical Geogra- density of 17-year periodical cicadas (Homoptera:
phy 17:62–76. Cicadidae). American Midland Naturalist 137:124–135.
LLOYD, M., AND R. KARBAN. 1983. Chorusing centers of RUBBO, M. J., J. J. COLE, AND J. M. KIESECKER. 2006. Terrestrial
periodical cicadas. Journal of the Kansas Entomological subsidies of organic carbon support net ecosystem
Society 56:299–304. production in temporary forest ponds: evidence from
LLOYD, M., AND J. WHITE. 1976. On the oviposition habits of an ecosystem experiment. Ecosystems 9:1170–1176.
13-year versus 17-year periodical cicadas of the same RUBBO, M. J., AND J. M. KIESECKER. 2004. Leaf litter composition
species. Journal of New York Entomology Society 84: and community structure: translating regional species
148–155. changes into local dynamics. Ecology 85:2519–2525.
MARSHALL, D. C. 2001. Ecological and population biology: STERNER, R. W., AND J. J. ELSER. 2002. Ecological stoichiometry:
periodical cicada (Homoptera: Cicadidae) life-cycle the biology of elements from molecules to the biosphere.
variations, the historical emergence record, and the Princeton University Press, Princeton, New Jersey.You can also read
